A computer-based method for reducing or eliminating baseline drift from a biological (bio) signal includes the steps of dividing the bio signal into a plurality of shorter signals having fixed time intervals, fitting a corresponding portion of a baseline function to a baseline of a respective one of each of the shorter signals, and subtracting the baseline function from the bio signal, resulting in a bio signal with a flat baseline.
|
1. A computer-based method for reducing baseline drift in a biological (bio) signal, the method comprising:
dividing the bio signal into a plurality of shorter signals having fixed time intervals;
fitting a corresponding portion of a baseline function to a baseline of a respective one of each of the shorter signals;
subtracting the baseline function from the bio signal, thereby flattening the baseline of the bio signal;
setting a total number of the time intervals to be equal to a total number of samples of the bio signal contained in a single cycle of a first frequency value associated with the baseline drift; and
repeating the steps of dividing, fitting, subtracting and setting, while substituting a second frequency value associated with the baseline drift for the first frequency value.
26. A computer-readable storage medium having stored thereon a series of instructions executable by a processor for removing noise from a digital signal, the instructions configured to cause the processor to perform the steps of:
dividing the bio signal into a plurality of shorter signals having fixed time intervals;
fitting a corresponding portion of a baseline function to a baseline of a respective one of each of the shorter signals;
subtracting the baseline function from the bio signal, thereby flattening the baseline of the bio signal;
setting a total number of the time intervals to be equal to a total number of samples of the bio signal contained in a single cycle of a first frequency value associated with the baseline drift; and
repeating the steps of dividing, fitting, subtracting and setting, while substituting a second frequency value associated with the baseline drift for the first frequency value.
12. A computer-based method for reducing baseline drift in a biological (bio) signal, the method comprising:
dividing the bio signal into a plurality of shorter signals having fixed time intervals;
fitting a baseline function to a baseline of each of the shorter signals;
subtracting the baseline function from the bio signal, thereby flattening the baseline of the bio signal;
setting a total number of the time intervals to be equal to a total number of samples of the bio signal contained in a single cycle of a first frequency value associated with the baseline drift; and
repeating the steps of dividing, fitting, subtracting and setting, while substituting a second frequency value associated with the baseline drift for the first frequency value, wherein the steps are repeated as a function of a comparison between a variation in a value of the baseline function to a resolution of an analog-to-digital converter used to generate the bio signal.
10. A computer-based method for reducing baseline drift in a biological (bio) signal, the method comprising:
dividing the bio signal into a plurality of shorter signals having fixed time intervals;
fitting a baseline function to a baseline of each of the shorter signals;
subtracting the baseline function from the bio signal, thereby flattening the baseline of the bio signal;
setting a total number of the time intervals to be equal to a total number of samples of the bio signal contained in a single cycle of a first frequency value associated with the baseline drift;
repeating the steps of dividing, fitting, subtracting and setting, while substituting a second frequency value associated with the baseline drift for the first frequency value, wherein the steps are repeated based on an evaluation of a flatness of the bio signal after the subtracting is performed; and
evaluating the flatness by fitting an evaluation function to the bio signal and examining a magnitude of the evaluation function.
14. A device for reducing baseline drift in a biological (bio) signal, the device being constructed and arranged to:
divide the bio signal into a plurality of shorter signals having fixed time intervals;
fit a corresponding portion of a baseline function to a baseline of a respective one of each of the shorter signals;
subtract the baseline function from the bio signal, thereby flattening the baseline of the bio signal; and
set a total number of the time intervals to be equal to a total number of samples of the bio signal contained in a single cycle of a first frequency value associated with the baseline drift, the device comprising:
a communications arrangement configured to receive the bio signal; and
a processor configured to divide the bio signal into a plurality of shorter signals having fixed time intervals, fit the corresponding portion of the baseline function to the baseline of the respective one of each of the shorter signals, subtract the baseline function from the bio signal, thereby flattening the baseline of the bio signal, and set the total number of the time intervals to be equal to the total number of samples of the bio signal contained in the single cycle of the first frequency value associated with the baseline drift; and
a memory including instructions configuring the processor to divide the bio signal into the plurality of shorter signals having fixed time intervals, fit the corresponding portion of the baseline function to the baseline of the respective one of each of the shorter signals, subtract the baseline function from the bio signal, thereby flattening the baseline of the bio signal, and set the total number of the time intervals to be equal to the total number of samples of the bio signal contained in the single cycle of the first frequency value associated with the baseline drift,
wherein the instructions direct the processor to:
repeat the steps of dividing, fitting, subtracting and setting, while substituting a second frequency value associated with the baseline drift for the first frequency value.
24. A device for reducing baseline drift in a biological (bio) signal, the device being constructed and arranged to:
divide the bio signal into a plurality of shorter signals having fixed time intervals;
fit a baseline function to a baseline of each of the shorter signals;
subtract the baseline function from the bio signal, thereby flattening the baseline of the bio signal; and
set a total number of the time intervals to be equal to a total number of samples of the bio signal contained in a single cycle of a first frequency value associated with the baseline drift, the device comprising:
a communications arrangement configured to receive the bio signal;
a processor configured to divide the bio signal into the plurality of shorter signals having fixed time intervals, fit the baseline function to the baseline of each of the shorter signals, subtract the baseline function from the bio signal, thereby flattening the baseline of the bio signal, and set the total number of the time intervals to be equal to the total number of samples of the bio signal contained in the single cycle of the first frequency value associated with the baseline drift; and
a memory including instructions configuring the processor to divide the bio signal into the plurality of shorter signals having fixed time intervals, fit the baseline function to the baseline of each of the shorter signals, subtract the baseline function from the bio signal, thereby flattening the baseline of the bio signal, and set the total number of the time intervals to be equal to the total number of samples of the bio signal contained in the single cycle of the first frequency value associated with the baseline drift,
wherein the processor is configured to repeat the steps of dividing, fitting, subtracting and setting, while substituting a second frequency value associated with the baseline drift for the first frequency value, and wherein the steps are repeated as a function of a comparison between a variation in a value of the baseline function to a resolution of an analog-to-digital converter used to generate the bio signal.
22. A device for reducing baseline drift in a biological (bio) signal, the device being constructed and arranged to:
divide the bio signal into a plurality of shorter signals having fixed time intervals;
fit a baseline function to a baseline of each of the shorter signals;
subtract the baseline function from the bio signal, thereby flattening the baseline of the bio signal; and
set a total number of the time intervals to be equal to a total number of samples of the bio signal contained in a single cycle of a first frequency value associated with the baseline drift, the device comprising:
a communications arrangement configured to receive the bio signal;
a processor configured to divide the bio signal into the plurality of shorter signals having fixed time intervals, fit the baseline function to the baseline of each of the shorter signals, subtract the baseline function from the bio signal, thereby flattening the baseline of the bio signal, and set the total number of the time intervals to be equal to the total number of samples of the bio signal contained in the single cycle of the first frequency value associated with the baseline drift; and
a memory including instructions configuring the processor to divide the bio signal into the plurality of shorter signals having fixed time intervals, fit the baseline function to the baseline of each of the shorter signals, subtract the baseline function from the bio signal, thereby flattening the baseline of the bio signal, and set the total number of the time intervals to be equal to the total number of samples of the bio signal contained in the single cycle of the first frequency value associated with the baseline drift,
wherein the processor is configured to:
repeat the steps of dividing, fitting, subtracting and setting, while substituting a second frequency value associated with the baseline drift for the first frequency value, wherein the steps are repeated based on an evaluation of a flatness of the bio signal after performing the subtracting, and
evaluate the flatness by fitting an evaluation function to the bio signal and examining a magnitude of the evaluation function.
2. The method of
3. The method of
4. The method of
5. The method of
6. The method of
8. The method of
9. The method of
11. The method of
repeating the steps when the magnitude exceeds a predetermined threshold value.
13. The method of
repeating the steps when the variation exceeds the resolution.
15. The device of
16. The device of
17. The device of
18. The device of
19. The device of
21. The device of
23. The device of
25. The device of
27. The computer-readable storage medium of
|
This invention relates to a method, a system and a device for processing electrocardiac signals, or other biological (bio) signals, to reduce or eliminate baseline drift.
An electrocardiogram (ECG or EKG), is a graphic produced by an electrocardiograph, which records the electrical activity (a signal) of the heart over time. Electrical waves cause the heart muscle to pump. These waves pass through the body and can be measured at electrodes (electrical contacts) attached to the skin. Electrodes on different sides of the heart measure the activity of different parts of the heart muscle. An EKG displays the voltage (a signal) between pairs of these electrodes, and the muscle activity that they measure, from different directions. This display indicates the overall rhythm of the heart, and weaknesses in different parts of the heart muscle. It is a way to measure and diagnose abnormal rhythms of the heart, particularly abnormal rhythms caused by damage to the conductive tissue that carries electrical signals, or abnormal rhythms caused by levels of salts, such as potassium, that are too high or low.
The ability to analyze an EKG signal and detect variances therein, allows for monitoring the physiological condition of a heart. For instance, accurate detection of variances in an EKG signal allows for the detection of heart events, such as heartbeat detection, arrhythmias, ischemias, and a myriad of other events. To detect variances in the EKG signal, it is necessary to minimize or eliminate noise, which also causes variances in an EKG signal, but does not correspond to a physiological event of the heart. Otherwise, the noise variance may be misconstrued as a heart event. In turn, this can lead to a potential misdiagnosis, false positive, missed event, or failure to detect other rhythms, among other undesirable results.
Baseline drift is a type of noise that causes signals, such as an EKG signal, to wander, i.e., drift in a linear or nonlinear fashion. Baseline drift may arise from any number of factors including, but not limited to, drift in electronic signal conditioning, thermal or mechanical stresses at the electrodes, and changes in operation condition, e.g., variations in ambient or body temperature, patient movement, etc.
Present techniques used to minimize baseline drift involve the use of filters. For example, one known method of removing baseline drift involves the use of a high-pass filter to filter out frequencies below a selected cutoff frequency. High pass filters, however, deviate from their ideal models, resulting in undesirable performance. In particular, high pass filters feature “roll-off”, which refers to imperfections in the signal response of a digital filter around a cut-off value.
Digital filters attempt to approximate the desired ideal response by increasing the length (order) of their impulse response. Because the digital filters must be causal, delay of the output is a necessary result. The main tradeoffs in digital filters include additional delay and increased numeric precision required as the order of the filter is increased. Therefore, the roll-off is unavoidable as well as a significant delay in the signal when close approximations are required.
Based on the above descriptions of conventional high pass filters, it can be seen that any realizable high pass filters cannot fully attenuate low frequency noise such as baseline drift in EKG signals (or other bio signals, such as from the brain). The technical problem of baseline drift and other such frequency based noise in a bio signal is potentially extremely serious. This is particularly so when experienced in the field of, for example, EKG signals since errors occurring in readings of an EKG signal can lead to erroneous deductions as to patient condition or required treatments.
According to a first aspect of the present invention, there is provided a computer-based method for reducing baseline drift in a biological (bio) signal, the method comprising: dividing the bio signal into a plurality of shorter signals having fixed time intervals; fitting a corresponding portion of a baseline function to a baseline of a respective one of each of the shorter signals; and subtracting the baseline function from the bio signal, thereby flattening the baseline of the bio signal.
In one embodiment the method is for eliminating baseline drift, resulting in a flat baseline.
According to a second aspect of the present invention, there is provided a device for reducing baseline drift in a biological (bio) signal, the device being constructed and arranged to: divide the bio signal into a plurality of shorter signals having fixed time intervals; fit a corresponding portion of a baseline function to a baseline of a respective one of each of the shorter signals; and subtract the baseline function from the bio signal, thereby flattening the baseline of the bio signal.
Preferably, the device completely eliminates baseline drift, resulting in a flat baseline.
In one embodiment, the device comprises: a communications arrangement configured to receive the bio signal; and a processor to divide the bio signal into a plurality of shorter signals having fixed time intervals, fit a corresponding portion of a baseline function to a baseline of a respective one of each of the shorter signals, and subtract the baseline function from the bio signal, resulting in a bio signal with a flat baseline. In other words, the device includes a processor that is configured to execute the steps on the bio signal so as to produce as a result a bio signal with a flattened baseline.
In one embodiment, the device includes a memory which itself includes instructions configuring the processor to divide the bio signal into a plurality of shorter signals having fixed time intervals, fit a corresponding portion of a baseline function to a baseline of a respective one of each of the shorter signals; and subtract the baseline function from the bio signal, thereby flattening the baseline of the bio signal.
According to a third aspect of the present invention, there is provided a computer-readable storage medium having stored thereon a series of instructions executable by a processor for removing noise from a digital signal, the instructions configured to cause the processor to perform the steps of: dividing the bio signal into a plurality of shorter signals having fixed time intervals; fitting a corresponding portion of a baseline function to a baseline of a respective one of each of the shorter signals; and subtracting the baseline function from the bio signal, thereby flattening the baseline of the bio signal.
Exemplary embodiments of the present invention will be described with reference to the removal of low frequency baseline drift from EKG signals. However, it will be appreciated that the exemplary embodiments of the present invention described below may be adapted towards the removal of any type of frequency based noise. Furthermore, the present invention need not be limited to bio signals, but may be applied to any digital signal.
In 310, a digitally sampled EKG signal may be received. The digital signal may correspond to an analog EKG signal recorded by a patient monitoring device after the analog signal has been passed through an analog-to-digital (A/D) converter. The digital signal may illustrate one or more heartbeats sensed by an electrode of the monitoring device.
Referring back to
In 314, the number of samples contained in a single cycle of the cutoff frequency may be determined. The determination may be performed as a function of a sampling frequency of the digital signal. For example the signal 20 was sampled at a frequency of 1,000 Hz so there are 1,000 samples every second. A cutoff frequency of 1 Hz would, accordingly, correspond to the number of samples contained in one second of data, e.g., 1,000 samples.
In 316, the digital signal may be divided into time intervals equal in length to the determined number of samples. The time intervals may comprise breakpoints for use in fitting a baseline function to the digital signal, as will be described below. If the signal 20 has a length of 100,000 samples, dividing the signal 20 into intervals of 1,000 samples would yield 100 breakpoints. Thus, the location of the breakpoints may be determined as a function of the sampling frequency of the digital signal and the frequencies of interest, e.g., frequencies that correspond to noise.
In 318, the digital signal may be fitted over each time interval to generate a baseline function that spans the entire length of the digital signal. The baseline function may be a piece-wise function comprising a plurality of smaller baseline functions corresponding to each time interval, and formed using cubic-spline interpolation. A cubic polynomial may be generated to fit the digital signal values between each breakpoint, yielding a baseline function that approximates the baseline drift. For each time interval, a cubic function is generated such that it passes through the signal values located at the beginning and end of the time interval, e.g., the breakpoints.
The exactness of the fit may depend on how many breakpoints are used. Therefore the value of the cutoff frequency selected may influence fit. Generally, the more breakpoints selected (i.e., higher cutoff frequency), the higher the degree of fit.
In 320, the baseline function may be subtracted from the digital signal to generate a flat-baselined EKG signal. Each value in the baseline function may be deducted from a corresponding digital signal value to generate the flat-baselined signal.
The fitting of the baseline function and the subsequent subtracting of the baseline function from the digital signal may be repeated at different frequencies within the range of frequencies associated with the noise source. Fitting may be repeated based on a determination of whether results of the signal analysis are satisfactory.
One measure of whether the results are satisfactory is to compare the variation in the baseline function to the maximum resolution of a digital representation of the baseline function. The resolution of an A/D converter unit determines the smallest error that can be represented by the A/D unit's output. As an illustrative example, if the gain is 6,400 A/D units per millivolt, then a change in the LSB corresponds to a change of 1/6,400 mV=156 microvolts. If the variation in the baseline function is less than 156 microvolts, then no further improvement may be possible given the A/D resolution. However, if the variation is larger than the magnitude of the LSB change, then the fitting may be repeated, e.g., repeating the cubic-spline interpolation on the new waveform. Each time the fitting is repeated, the variation in the baseline drift may be reduced. In this manner, the fitting may be repeated until the variation is less than or equal to the magnitude of the LSB change. Accordingly, in 322, a determination may be performed whether the variation in the baseline function exceeds the A/D unit's resolution.
If the variation exceeds the A/D unit's resolution, then the method 300 may proceed to 324, where a higher frequency is selected from the frequency range corresponding to the baseline drift. The method 300 then returns to 314.
The fitting may also be repeated based on an evaluation of a flatness of the flat-baselined signal. Flatness may be evaluated by fitting an evaluation function to the flat-baselined signal, e.g., using cubic-spline interpolation or least squares fitting, and viewing the magnitude of the evaluation function.
If the magnitude exceeds the threshold value, then the method 300 may proceed to 324 before returning to 314.
If the magnitude does not exceed the threshold value, then the method 300 may proceed to 328, where a signal analysis may be performed on the flat-baselined signal. Clinical points of interest may be determined by, for example, determining the locations of QRS complexes in the flat-baselined signal. The ORS complexes correspond to individual heartbeats and may be found by locating large spikes in the flat-baselined signal, which may be easier to do now that the digital signal has been de-noised. Detection of other points of interest may also be easier using the flat-baselined signal.
Although the method 300 was described with reference to automatic repeated fitting, the repeated fitting may also be performed manually, e.g., by the user manually selecting new cutoff frequencies based on a manual evaluation of flatness. In one embodiment, the user may select random values within the frequency range. In another embodiment, the user may choose to gradually increase the cutoff frequency at each iteration. Cutoff frequency selection may also be selected (either manually or automatically) by, for example, increasing the cutoff frequency by a predetermined amount each time the fitting is repeated (e.g., fixed step or linear increases). In this manner, baseline drift associated with a variety of frequencies may be removed.
It will be also be appreciated that other fitting techniques may be used in addition to cubic-spline interpolation. For example, other types of polynomial interpolation, e.g., quadratic may be used. Polynomial interpolation generally yields a baseline function that passes through each signal value located at the breakpoints. Other fitting techniques may not require such an exact fitting. For example, in another embodiment, a least squares fit may be used to generate a baseline function that approximates the baseline drift without intersecting every signal value at the breakpoints.
The memory may be a computer-readable storage medium that includes any type of readable or writable memory, including RAM, ROM, flash memory, an optical or electromagnetic drive, a compact disc, etc. In addition to storing the instructions 122, the memory may also include data 124 used in performing the method 300. For example, the data 124 may include digital values, e.g., x-y coordinates, corresponding to the digital signal. The data 124 may also include the cutoff frequency, the range of frequency values, the baseline function and the flat-baselined signal.
The communications arrangement 130 may include any combination of hardware or software components for communicating with a data source such as a data collection device 150 or a computing device 160. The data source may be configured to collect the digital signal and transmit it to the device 100 via the communications arrangement 130. The communications arrangement 130 may be in wired and/or wireless communication with the data source. For example, the communications arrangement 130 may wirelessly communicate with the device 150. The communications arrangement 130 may also be in wired communication with the device 160 via a network 162, which may be a local area network, a wide area network, a telephone network, the Internet, etc.
Data may be collected and transmitted to the device 100 in substantially real time. For example, the device 150 may include sensor electrodes for generating EKG signals. An analog EKG signal may be processed and converted to a digital signal, then transmitted. Data collection may also be done any time after the analog signal is recorded. For example, analog or digital signals may be stored in a database on the device 160, batch processed, and transmitted together. The device 160 may, similar to the device 150, include sensors for measuring EKG signals. Alternatively, the device 160 may be configured to communicate with an external sensing device. In further embodiments, data may be transmitted to the device 160 via manual input, a storage device such as a CD-ROM, or any other input arrangement.
It will be appreciated that the example systems, devices and methods described above may be integrated into a system for monitoring patient health. An example of a system which may be suitable for use with the present invention is described in U.S. patent application Ser. No. 11/938,409, Method and System for Active Patient Management, which describes an active patient management system for monitoring patient EKG signals along with other health indicators. In one embodiment, the method 300 may be implemented in the active patient management system as a software program stored in a computer readable medium such as hard drive memory, flash memory, floppy disk memory, optically-encoded memory (e.g., a compact disk, DVD-ROM, DVD±R, CD-ROM, CD±R, holographic disk), a thermomechanical memory (e.g., scanning-probe-based data-storage), or any type of machine-readable (e.g., computer-readable) storage medium.
The example systems and methods of the present invention described above have been shown to reduce or eliminate low frequency baseline drift in EKG signals. It will be appreciated that the systems and methods of the present invention may also be used to reduce or eliminate high frequency noise in any type of digital signal. In another embodiment, high frequency cutoff values may be used to remove high frequency noise in EKG signals. In this manner, both high and low frequency noise may be reduced or eliminated.
In the preceding specification, the present invention has been described with reference to specific example embodiments thereof. It will, however, be evident that various modifications and changes may be made thereunto without departing from the broader spirit and scope of the present invention as set forth in the claims that follow. The specification and drawings are accordingly to be regarded in an illustrative rather than restrictive sense.
Patent | Priority | Assignee | Title |
Patent | Priority | Assignee | Title |
5704365, | Nov 14 1994 | SPACELABS HEALTHCARE, INC | Using related signals to reduce ECG noise |
5713367, | Jan 26 1994 | SPACELABS HEALTHCARE, INC | Measuring and assessing cardiac electrical stability |
6470320, | Mar 17 1997 | BOARD OF REGENTS OF THE UNIVERSITY OF OKLAHOMA, THE | Digital disease management system |
6533724, | Apr 26 2001 | ABIOMED, INC | Decision analysis system and method for evaluating patient candidacy for a therapeutic procedure |
20020099686, | |||
20030101076, | |||
20040015337, | |||
20040078232, | |||
20040103001, | |||
20040236188, | |||
20050119534, | |||
20060025931, | |||
20060173663, | |||
20060224416, |
Executed on | Assignor | Assignee | Conveyance | Frame | Reel | Doc |
Sep 25 2008 | Air Products and Chemicals, Inc. | (assignment on the face of the patent) | / | |||
Sep 25 2008 | TOTH, MICHAEL S | Air Products and Chemicals, Inc | ASSIGNMENT OF ASSIGNORS INTEREST SEE DOCUMENT FOR DETAILS | 021835 | /0124 |
Date | Maintenance Fee Events |
Jun 10 2016 | REM: Maintenance Fee Reminder Mailed. |
Oct 30 2016 | EXP: Patent Expired for Failure to Pay Maintenance Fees. |
Date | Maintenance Schedule |
Oct 30 2015 | 4 years fee payment window open |
Apr 30 2016 | 6 months grace period start (w surcharge) |
Oct 30 2016 | patent expiry (for year 4) |
Oct 30 2018 | 2 years to revive unintentionally abandoned end. (for year 4) |
Oct 30 2019 | 8 years fee payment window open |
Apr 30 2020 | 6 months grace period start (w surcharge) |
Oct 30 2020 | patent expiry (for year 8) |
Oct 30 2022 | 2 years to revive unintentionally abandoned end. (for year 8) |
Oct 30 2023 | 12 years fee payment window open |
Apr 30 2024 | 6 months grace period start (w surcharge) |
Oct 30 2024 | patent expiry (for year 12) |
Oct 30 2026 | 2 years to revive unintentionally abandoned end. (for year 12) |